Letter Regarding Article by Diwan et al, “Doppler Estimation of Left Ventricular Filling Pressures in Patients With Mitral Valve Disease”
To the Editor:
In the article on Doppler estimation of left ventricular filling pressures in patients with mitral valve disease, Diwan et al1 measured the time interval between the onset of early diastolic mitral inflow (E) and annular early diastolic velocity (Ea) by tissue Doppler (TDI), expressed as TE–Ea. We have concerns about the accuracy of this measurement. To avoid large differences in the diastolic time intervals with different R-R cycle lengths, they said that identical R-R intervals (≤5 ms) were chosen for timing the onset of mitral E and the onset of mitral annular Ea. But even in Figure 1 of the article by Diwan et al,1 after lining the vertical lines at the peak of the R wave, we found that the R waves of the ECGs did not coincide with each other and the cardiac cycle length between 2 examinations differed by more than 6% of the R-R interval. In a previous article by Sohn et al,2 this approach was criticized, as with many of the cases in their study, it was impossible to obtain cardiac cycles with the same R-R intervals.
Assuming the sweep speed and scale of 2 figures to be same, we moved the lower figure slightly to the left to synchronize the first R wave, after which the onset points of the mitral inflow and annular movement then coincided exactly, ie, TE–Ea was almost zero. Sohn et al2 also criticized the clinical validity of the formula isovolumic relaxation time (IVRT)/TE–Ea as not being relevant, as in the majority of cases in their study, TE–Ea was zero.
Therefore, the Figure chosen in support of this study caused some questions about the validity of this measurement. Considering that small differences in the TE–Ea measurement can make the value of IVRT/TE–Ea value differ 2- or 3-fold, we suggest that stricter image analysis with R wave synchronization should be performed for verification of the time intervals data.
We appreciate the interest of Park et al in our article.1 Although the difference in the R-R interval for the mitral inflow and tissue Doppler (TD) cycles is 5% (with the average of 5 ms difference for 3 cardiac cycles), the duration is longer for the R-R interval in the mitral inflow cycle. Likewise, the mitral regurgitation duration by CW Doppler (which is equal to the sum of the isovolumic contraction period, systolic ejection period, and isovolumic relaxation period) is longer in the mitral cycle, resulting in a later onset of mitral inflow (E) in comparison with the R-R interval in the TD recording, ie, moving mitral E closer to annular early diastolic volume (Ea) in the TD recording. Correcting the time from R peak to mitral onset for the R-R duration in the TD cycle results in an even longer TE-Ea.
The peak of the R waves in the Figure are reasonably well aligned. We are therefore surprised that the “synchronizing” attempt by Park et al resulted in abolishing a time difference of 56 ms between mitral E and annular Ea (as implied in their letter). As for the ratio of isovolumic relaxation time (IVRT) to TE-Ea, we agree and reiterate what was previously stated in our article1; namely, the ratio is not applicable in situations with simultaneous onset of mitral E and annular Ea. Nevertheless, in hearts with normal left ventricular relaxation, where TE-Ea can be zero, the presence of a normal or prolonged IVRT would be indicative of a corresponding normal or reduced left atrial pressure.
Finally, more important than arguing R wave “synchronization,” we are in strong agreement with efforts to reach reliable measurements from a single cardiac cycle, as recommended by Oh and Tajik2 and Oh.3 Such an approach would reduce measurement errors and measurement time. Furthermore, the ratio of IVRT to tau can still be used for the estimation of LV filling pressures.